Archive notice: This article was originally published on August 2, 2005. Links and embedded videos are preserved as part of the historical record.
With the help of software from MIT (Massachussets Institute of Technology), minimal movements in videos can be magnified: After the video has been loaded into the software, it searches the image for trackable points, whose movement is then analysed.
In this way, the moving parts are filtered out of the video. The moving pixels are combined into regions and can be supplemented by the user. The regions serve as masks to divide the video into different layers.
These layers are placed on top of one another, and the movement of the individual pixels is amplified. The results are astonishing – they can be viewed on the project’s homepage.
Further Links:
- Homepage of the “Motion Magnification” project
